Thermodynamic and material properties of reversible cluster formation: application to concentrated protein solutions

Wagner, Norman J.Liu, YunColloidal particles and ``soft matter'' materials are essential parts of many consumer goods we utilize in all facets of life. The utility of these materials stems from their chemical and structural versatility, ease of customization, and response to external stimuli. As a result, soft materials are ubiquitous in industries such as electronics, performance composites, personal care, cosmetics, paints and pharmaceuticals. Engineering these materials relies on a fundamental understanding of their phase behavior and structure-property relationships that dictate various microstructural states for particular applications. Regardless of the chemistry of various colloids, whether they are particles, polymers, micelles, or proteins, the combination of interaction forces (e.g., dispersion, hydrogen bonding, hydrophobic, electrostatic) can be represented as an effective potential energy field, or interaction potential, which determines the diversity of equilibrium and non-equilibrium states. This thesis examines the influence of a combination of a short-range attraction (SA) and long-range repulsion (LR), ubiquitous to protein systems, on the phase behavior and associated material properties. The competition of attractive and repulsive forces is unique in its ability to produce self-assembled clusters with a preferred size. This type of cluster has been observed in a variety of materials. For example, the biopharmaceutical industry is concerned with large solution viscosities in highly concentrated therapeutic formulations that are hypothesized to arise from cluster formation. In general, SALR (or competing) interactions may have a significant impact on material properties stemming from the large diversity of states they may form. Thus, theoretical calculations and experimental observation of thermodynamic and material properties (e.g., viscosity) are necessary to more accurately understand the phase behavior of SALR systems. Characteristic features of cluster formation in SALR systems are investigated and identified using scattering techniques. In contrast to early experimental studies, the presence of a unique intermediate range order (IRO) peak in small angle scattering patterns is found not to be an experimental representation of cluster formation. Using the particle level details provided by Monte Carlo (MC) simulations, several microstructures are studied and distinguished using well defined definitions according to the cluster size distribution. The contribution to the scattering intensity is decomposed into correlations between monomers and clusters. The results indicate that a significant contribution to the IRO peak is from monomers in each type of microstructure. Some specific properties of this peak are found to be useful identifiers of clustered states in the one phase region. The widely used extended law of corresponding states for colloidal systems with only an SA interaction is extended in this thesis to systems with competing interactions. A generalized phase diagram for systems with isotropic SALR potentials has been identified to distinguish different liquid states including clustered fluid states. Fluids with a preferred cluster size driven by SALR interactions are identified using MC simulations, and are found to form exclusively within the two-phase region of a purely attractive reference system. The additional repulsion frustrates phase separation, driving particle localization on intermediate range order (IRO) length scales. Multiple potential forms and interaction parameter sets are investigated and demonstrate identical behavior, ensuring that this generalized phase diagram is a generic feature of systems with competing interactions. This phase diagram serves as an effective and efficient method of identifying cluster formation for the community. A model protein solution is used to experimentally study the relation between cluster formation and dramatic increases in viscosity. Using neutron scattering techniques, the structure and dynamics of this nearly isotropic system under concentrated conditions are accurately quantified. Interaction parameters are extracted by fitting neutron scattering data with a thermodynamically self-consistent integral equation theory. The extent of cluster formation is identified by mapping these states onto the new generalized phase diagram and generating structures using MC simulations. Unique glassy-like behavior and large viscosities are shown to arise from structures with IRO. A new viscosity model is developed for SALR systems to capture the additional impact of cluster interactions. These methods are applied to understand the anomalously large viscosities sometimes apparent in concentrated monoclonal antibody solutions, which are used as biopharmaceutical therapeutics. The formation of dynamic clusters had been hypothesized to be the underlying driving force behind the viscosity increase at higher concentration. Three different mAbs are studied over a wide range of commercially relevant formulation conditions to distinguish the thermodynamic properties of viscous solutions. The formation of long-lived dimers is consistently found to cause the larger viscosities. Differences in the magnitude depend on the subsequent association of these small clusters into supramolecular structures. Model systems could be used as a foundation to semi-quantitatively characterize the extent of cluster formation and its influence on viscosity.University of Delaware, Department of Chemical and Biomolecular EngineeringPh.D